AI chatbots and digital coaching tools are reaching people at nearly every stage of the justice system—from prison tablets to probation apps. Reliance on them is outpacing the standards that should govern their use.
The key question isn’t whether a machine can understand someone as well as a person can. It’s whether the person on the other end can question what it tells them, and knows who else is listening.
Our brief suggests practical standards for the agencies buying these tools, the vendors building them, and the human clinicians already doing this work.
